Diagnostic Test Explorer

Disease severity

How much lower do people with the disease score on the diagnostic test, on average? This also represents how precisely the test can measure disease characteristics.

Mean difference20

060

Diagnostic cutoff

Scores at or below the cutoff are a positive result (conclude that the patient has the disease).

Highest balanced accuracy with cutoff —

Cutoff score90

0200

Prevalence

The percentage of people in the population being tested who have the disease.

Prevalence50%

0%100%

Healthy M=100 SD=15 Disease M=80 SD=15 Cutoff

Test score

ROC curve Current cutoff

1 – Specificity (false positive rate)

Diagnostic metrics

Sensitivity

—

TP / (TP+FN)

Rate of correct positive results among those with disease

False negative rate = —

SnNout

High sensitivity (Sn), when negative (N), rules out the disease (out).

Specificity

—

TN / (TN+FP)

Rate of correct negative results among those without disease

False positive rate = —

SpPin

High specificity (Sp), when positive (P), rules in the disease (in).

Balanced Accuracy

—

(Sens + Spec) / 2

Average accuracy across both groups; useful when classes are unequal

NPV

—

TN / (TN+FN)

Probability the patient is disease-free given a negative test

PPV

—

TP / (TP+FP)

Probability the patient has disease given a positive test

LR−

—

(1−Sens) / Spec

How much a negative result decreases the odds of disease

LR+

—

Sens / (1−Spec)

How much a positive result increases the odds of disease

Results per 1,000 patients

	Disease+	Disease−	Total
Test+	—	—	—
Test−	—	—	—
Total	—	—	1000

Disease+, Test+ (TP) Disease+, Test− (FN) Disease−, Test+ (FP) Disease−, Test− (TN)

Each dot = 10 patients

Summary